Source Room / White paper
One Hotel, Seven Versions: sources, dates and limits
Before citing a finding about a room, a booking condition or an AI answer, check which source was examined, when it was recorded and what remained unverified.
This note explains the evidence behind One Hotel, Seven Versions, how to interpret its labels and how to submit a correction.
What the study examined
The study follows 16 traveller-critical facts across seven representation roles for Hôtel Mont-Blanc Chamonix:
- The hotel’s own pages.
- Its direct booking engine.
- Structured data.
- Google and local-search surfaces.
- Online travel agencies and reviews.
- Destination and editorial sources.
- AI synthesis.
This is an independent Tamaga public-source study. It examines how information changes between representations. The case concerns a real property, separate from the fictional Tamaga Hotel demonstration.
To check a particular finding, start with the detailed observations in Appendix A, then consult the corresponding numbered references.
Which sources and dates matter?
The material comes from different periods. An observation date, a source’s publication date and a research edition date describe different things.
| Material | Recorded date | How to interpret it |
|---|---|---|
| Gemini report reproduced in the cited research | 20 December 2025 | A historical AI answer, rather than a fresh test of current responses. |
| Primary public-source review | July 2026 | The principal review period for the hotel’s public representations. |
| Fixed PDF edition | August 2026 | An edition identifier; the month does not establish an exact release day or a new observation date. |
| Research presentation and disclosure alignment | 3 November 2026 | An editorial update that did not change the case’s observation dates. |
The structured-data analysis concerns a proposal published by Jean-Claude Morand and Roland Schegg. Its deployment on the hotel’s website has not been verified. The research-lineage statement explains how their work informed Tamaga’s independent study.
Tamaga’s illustrative JSON-LD pattern in Appendix B demonstrates modelling choices. It is not a verified production implementation.
Read the two kinds of labels correctly
The paper’s evidence labels describe the basis for a claim:
| Label | Meaning |
|---|---|
| Established | Supported by official specifications, stable documentation or strong peer-reviewed evidence. |
| Observed | Supported by a documented dataset or public-source review, without necessarily establishing causation or broader applicability. |
| Illustrative | A case observation used to explain a mechanism, without estimating how common it is. |
| Unknown | A relationship for which the available public evidence is insufficient. |
The heatmap uses a separate set of representation statuses: precise or confirmed; partial or generic; contradicted or erroneous; and not stated or not observable.
Those statuses describe the material examined within the review’s limits. They are not hotel-quality scores or current confirmations of availability, conditions or service.
What the case can establish
The case illustrates how consequential information can change, lose a qualification or conflict across representations. It provides a method for examining those differences.
Its limits remain important:
- One property cannot establish sector-wide prevalence.
- The review did not inspect private systems, internal analytics, contracts or non-public operating data.
- The direct booking engine was not fully observable through the text-only review. Unobserved information cannot be treated as proof that the engine lacks it.
- No live, date-specific comparison transaction established whether a direct offer was cheaper, more flexible or available for the recurring scenario.
- The study does not assess hospitality quality, commercial performance or legal compliance.
Prices, policies, availability, booking interfaces and AI responses may have changed since the recorded observations.
Cite an edition or send a correction
Choose the version you actually consulted:
- Maintained web paper: includes later clarifications. Record the URL and your access date when citing it.
- August 2026 PDF: preserves the fixed edition. Identify the edition and relevant page.
- Research overview: provides a shorter introduction. Use the full paper for detailed findings and references.
For a correction, send the relevant passage or fact, its source, the issue identified and supporting evidence with its date. A later hotel update may establish that circumstances changed without showing that the original observation was wrong.
The Source Room correction policy explains the wider approach.